Making sense of measurement: What does good look like?

In this second commentary from the WFA Workshop on Programmatic Proficiency held in Singapore this April, Infectious Media’s Global Client Analytics Lead, Chris Cox offers his thoughts on how, for marketers to truly make sense of measurement, they need to tackle technology, data and analytics head on. All at the same time.

Marketers who get measurement wrong are not thinking big enough

The way some brands are trying to solve measurement is like six blind men trying to identify an elephant, lots of teams with only part of the problem all working in isolation. Previously, marketers could rely on their agency to provide perspective, but as marketers are increasingly bringing these challenges in-house, how can they step back to see the whole issue and address it in a joined-up way?

As marketers are increasingly taking the digital advertising reigns away from agencies, they are getting to grips with some knottier problems themselves, and there is none so complex as measurement. For many, measurement is in crisis, as developments in the industry are making good measurement harder and harder. Why is good quality measurement so difficult to achieve?

It’s because measurement is the result of a process which is more complex than ever, and one which relies on bringing together parts of the business which historically haven’t needed to operate as one unit. It means bringing together technology specialists, data engineers, analysts, data scientists and marketers. These are disparate parts of the business and often have no claim on each other’s time. So, they make decisions in isolation, leading to dislocated implementation.

For brands that work closely with an agency this problem can be somewhat simplified, since they can vet an agency ahead of time on their ability to work across these different disciplines. However, for brands trying to in-house this problem is significantly more acute, since it relies on changing business culture, and not just amongst the teams who need to do the work.

What is good measurement?

With such an intractable problem, the solution is often to measure what is easy. Clicks are a great example, both as a measure, and in attribution, where last click is still seen as a viable metric. However, it has been consistently proven that clicking audiences don’t correlate to buying audiences, and in fact most behave in the exact opposite way.

As a first step, good measurement relies on reliably identifying audiences. For example, you want a good sense of who did or didn’t see a piece of creative for incremental measurement to work, and you need to see consumers over time to understand attribution.

How to move forward

Technology, data and analytics are the three moving parts that need to be joined together to solve measurement. Without this, you could spend months bringing the right data into the business only to find you don’t have underlying technology that can efficiently handle the scale, or the reporting tools that can seamlessly integrate with that data.

Think about the tech – thoroughly, and throughout the organization

Developing a measurement strategy starts with a consideration of the technology needed. As technology suffers from the contradiction of being both rapidly evolving whilst relying on a long-term commitment to see benefit.

Getting senior stakeholders involved at this stage is critical, a business strategy that includes tech implementation needs high level buy in to ensure continuity over time. This project cannot be subject to career cycles and people coming and going from the business. Ideally, the CTO and CMO should be in alignment about what each is trying to achieve, and that vision should set the stage for all implementation.

Making measurement goals make sense

Measurement goals must be understood, articulated and be realistic. It isn’t realistic to go from a standing start to a large-scale attribution model built in-house in six months. It would be more achievable to have a strong base econometric model in that timeline, which would enable decision making and can help set the scene for later work. Marketers need to set achievable goals in the direction where the most meaningful gains can be made. Trying to skip to the end of the journey in measurement will always end in failure.

Humans at the heart of it

Finally, the business needs to bring together a coherent team which is united behind the measurement mission, responsible to each other and encompassing the various disciplines needed to solve these challenges. But the good news is this work doesn’t only benefit one area, fixing these problems will have positive ramifications beyond measurement.

Having scalable, efficient data storage and access is a universal good for a business. Similarly, with cross-team training and development plans, as these allow each part of the business to better understand both what is being attempted, and its utility.

But ultimately, as marketers seek the transparency and control of in-house ownership of measurement, this will require their business to be able to focus on the long-term, and often at a larger scale than is initially realised. To get to effective measurement brands need to develop a strong sense of what is possible and prepare themselves to accept that development comes with both successes and failures.

The views expressed in this opinion article belong to Chris Cox, Global Client Analytics Lead, Infectious Media.

For information about WFA's activities in APAC, please reach out to Ranji David at r.david@wfanet.org